Reserved topic scholarships

Department of Information Engineering and Computer Science

A1 - Enhancing Human-Object Interaction Understanding with large multimodal model (1 grant)

This PhD project aims to develop advanced models for human-object interaction using large vision and language models. By integrating multi-modal data, including visual and contextual information, the research seeks to improve the recognition, prediction, and personalization of interactions in dynamic environments.

Contact: Elisa Ricci e.ricci [at] unitn.it

​​​​A2 - Interpretable Autonomous Driving (project UE H2020 AI4MEDIA Sebe, G.A. 951911, CUP  E64I20000480006) (1 grant)

Learning-based end-to-end autonomous driving approaches have achieved considerable progress in recent years and the mainstream follows the imitation learning pattern, where the agent is encouraged to clone the expert's behavior. However, there exist clear weaknesses, including poor interpretability, strong dependency on the pre-trained experts leveraging privileged information, and lack of capacity for online learning. This project aims to develop a novel model-based reinforcement learning approach for interpretable autonomous driving. This envisioned approach involves action-conditioned future prediction and consequence evaluation of actions, which identifies the optimal one from candidates.
The proposed framework has the potential to combine with multi-modal learning (e.g., lidar, language modeling) and online tuning for domain adaptation through transfer learning and continual learning.

Contact: Niculae Sebe niculae.sebe [at] unitn.it

​​​​​​B1 - Human-aware motion planning for collaborative robotics (project UE HE MAGICIAN Fontanelli; G.A. 101120731, CUP: E63C23000730006) (1 grant)

The candidate will have to study solutions for human-aware motion planning in the context of collaborative industrial applications. After surveying the state of the art s/he will develop a motion prediction algorithm able to execute in real-time along with a planning algorithm that will strike a good compromise between productivity, safety and well being of the human operators.

Contact: Luigi Palopoli luigi.palopoli [at] unitn.it Daniele Fontanelli daniele.fontanelli [at] unitn.it

​​​​​​D2 - Artificial intelligence and machine learning for the automatic analysis of planetary radar data (project: ASI JUICE RIME 2021; CUP: F65F21000950005) (1 grant)

Space missions for the exploration of planets and celestial bodies in the Solar system are very important for the huge scientific and technological return associated with them. Some of the most challenging science objectives of these missions are related to the analysis of sub-surface processes that are of crucial importance for many different implications they have. 
The research activities of this grant are related to planetary radar sounders, which are instruments for the study of the subsurface of the planets of the Solar system. These radars operate from satellite platforms and acquire images related to the subsurface of celestial bodies that can results in groundbreaking science results. The activity will be focused on the development of methodologies for the processing and the automatic analysis of these data. Special emphasis will be given to methodologies that exploit the most recent developments in the framework of machine learning and signal processing.
Research will be developed at the Remote Sensing Laboratory (https://rslab.disi.unitn.it/) and will be related to the Radar for Icy Moon Exploration (RIME) on board of the JUpiter ICy moons Explorer (JUICE) of the European Space Agency (see https://sci.esa.int/web/juice for more details on the mission).

Contact: Lorenzo Bruzzone lorenzo.bruzzone [at] unitn.it

​​​​​​D3 - Deep learning for the automatic analysis of satellite remote sensing images (project: ESA HRLC_CCI+ 2023 - Bruzzone; CUP E63C23002010006) (1 grant)

The automatic analysis of image time series acquired by Earth Observation satellites is crucial for many different applications (e.g., precision farming, forestry, analysis of urban areas, monitoring of natural disasters). One of the most important applications is related to the monitoring of the land cover in relation to climate changes. In this context, even if artificial intelligence and machine learning have been widely used for the automatic analysis of remote sensing data, there are still many methodological challenges and application issues that should be addressed in the analysis of long time series of satellite data. 
The research activities of this grant are related to the development of novel methodologies based on deep learning for the automatic analysis of time series of optical (multispectral) images acquired by Earth Observation satellites. The research will be focused on the problems of the automatic classification (sematic segmentation) and change detection in multispectral images acquired by different satellite systems. The activity will consider methodological research (for the development of novel methods) and the related application to real world scenarios.
Research will be developed at the Remote Sensing Laboratory (https://rslab.disi.unitn.it/) and will be linked to the High Resolution Land Cover project developed under the leadership of RSLab in the framework of the Climate Change Initiative of the European Space Agency (https://climate.esa.int/en/projects/high-resolution-land-cover/).

Contact: Lorenzo Bruzzone lorenzo.bruzzone [at] unitn.it

Eurac Research

A8 - AI-based solutions to enable high photovoltaic integration in future electricity grids(1 grant)  

Artificial Intelligence (AI) solutions can support the energy transition by facilitation high photovoltaic integration into the grid. The installation of intelligent electronic devices through the electricity grid allows us to collect a large amount of data not yet completely exploited for planning and real-time operations. Through advanced data analytics and machine learning will be possible to accurately predict photovoltaic power generation, enabling better planning, resource allocation and flexibility potential. These techniques can optimize grid operations by balancing supply and demand (also given by non-linear loads e.g. electric vehicles and heat pumps), improving energy storage management, and reducing reliance on fossil fuels. AI can also engage and educate end users, providing personalized suggestions on energy use, thus giving them a central role in the energy transition and encouraging active participation.
Throughout the duration of the PhD programme, the scholarship holder is required to spend at least 75% of his/her time at Eurac Research, in accordance with the study plan/tutorial schedule. This requirement applies even if the time spent at the institution is not continuous.

.Contact: Grazia Barchi grazia.barchi [at] eurac.edu

Fondazione Bruno Kessler (FBK)

A3 - Bringing Machine Learning and Inference to Resource-Limited Networked Embedded Systems(1 grant)

The Internet of Things (IoT) paradigm proposes massive increases in multimodal data creation on tiny devices located at the far edges of the computing infrastructure. Despite the resource constraints of these devices, communication constraints require the data they produce to be locally processed before transmission. Further, due to the increasing complexity of the data generated, machine learning techniques are being successfully applied at the far edge through so-called TinyML. While most TinyML techniques focus on the individual device, increasingly these devices are both networked together and/or connected to larger cloud or swarm based infrastructures, introducing challenges for managing communication and coordination, both in data collection and usage as well as regarding elements of distributed training and continuous learning. Research into these challenges requires innovation in i) applying novel techniques such as distillation, hardware aware scaling and neural architecture search to implement orchestratable TinyML algorithms, ii) innovative communication, including optimizing the low power communication network connecting edge devices to the infrastructure, and iii) orchestration techniques required for ML-enhanced edge devices to participate in complex distributed applications formed of heterogeneous devices. This interdisciplinary research at the intersection of Artificial Intelligence, Embedded Systems, Distributed Computing, and Low-power Hardware will take into account the candidate's profile and interests, contributing to the development of innovative solutions for real-world challenges. The candidate will have the opportunity to work with cutting-edge technology, gain valuable experience in interdisciplinary collaboration, and make significant contributions to the field of machine learning at the very edge.

Contact: Amy Lynn Murphy murphy [at] fbk.eu Elisabetta Farella efarella [at] fbk.eu

A4 - Cooperative Large Language Models (1 grant)

Nowadays, foundation models have shown remarkable capabilities in generating text, images and videos, rich world knowledge, and some complex “reasoning” skills. However, these models are still passive models (e.g., action-oriented aspects of intelligence are not leveraged), static, data- and computation-expensive, they often confabulate, and are difficult to align with human values.
This PhD project aims at making multimodal foundation models capable of interactions with other agents (e.g., enabling cooperative capabilities) and of interactions grounded in the world, enabling counterfactual reasoning and causality abilities in foundation models as well as enforcing their alignment with human intentions and values. The PhD candidate is required to have previous experience in working with deep learning algorithms, a strong interest on transformers’ architecture, graph neural networks, multimodality, and cooperative and embodied AI. The selected student will be able to collaborate with the ELLIS network and being part of the ELLIS PhD program (if selected) as well as with top universities and research centers such as MIT, Max Planck, etc.

Contact: Bruno Lepri lepri [at] fbk.eu

A5 - Neural Language Models for crisis management communication (1 grant)

The continuous release of new and increasingly powerful language models is opening possibilities to address applicative scenarios that were not even imaginable a few years ago. In particular, the goal of this PhD is to improve strategic and tactical communication activities during crisis events, through the use of advanced natural language generation and persuasive communication techniques. More specifically, the idea is to directly intervene with textual responses and narratives that are meant to address the public during and after crisis events and natural disasters. To this end the candidate should focus on all those aspects of the LLM, such as decoding strategies, knowledge guided generation, data quality, knowledge distillation, reinforcement learning from human feedback -just to mention a few- that can help in improving the models, especially for better factuality, reducing hallucination and increasing coherence among messages while assisting professionals in crisis management.

Contact: Marco Guerini m.guerini [at] fbk.eu

A6 - Speech Translation in the LLM Era (1 grant)

The advent of foundation models such as large language models (LLMs) has led to impressive advancements in all areas of natural language processing. Exploiting their capabilities for many downstream applications has therefore emerged as an interesting research direction, as well as extending them to other modalities. Toward this aim, researchers have started investigating the combination of speech encoders and LLMs. Despite the potential of this emerging approach, systematic studies are still required to isolate its strengths and weaknesses compared to traditional systems, as well as to identify the most effective strategies for specific applications, spanning from architectural choices to selecting the optimal data and tasks for training the adapter between the speech encoder and the LLM. In this scenario, the PhD candidate will study existing solutions, identify their weaknesses, and develop innovative approaches to overcome them, contributing to the advancement of the field, either on a theoretical basis, with an application-oriented focus, or both.

Contact: Matteo Negri negri [at] fbk.eu Luisa Bentivogli bentivo [at] fbk.eu

A7 - Modeling and simulation of Urban Digital Twins (1 grant)

An Urban Digital Twin is a dynamic digital model of a city, fed by data collected from the city itself, and capable of faithfully reproducing the city behaviour through the use of advanced modelling, data science, and AI techniques. One of the key foreseen applications of Urban Digital Twins is to predict the evolution of the city, including the effects and impacts of external changes (eg, climate change) and internal processes (eg, urban transition policities and incentives). A challenge for the development of Urban Digital Twins is that cities are systems-of-systems, with complex interactions between physical, organisational, and social dimensions, as well as between the physical world and the digital world. Novel modeling and simulation techniques are necessary to develop Urban Digital Twins able to manage this complexity and produce reliable predictions. 
The candidate will be requested to work in this research area, and in particular to work on advanced agent-based modeling and simulation frameworks for Urban Digital Twins and to their validation on real scenarios concerning the adoption of Digital Twin by Italian cities. Besides the requirements established by the rules of the ICT school, preferential characteristics for candidates for this scholarship are: 
• master's degree in Computer/Data Science, Mathematics, Phisics, Electrical Engineering, Communication Engineering, or equivalents; 
• knowledge in artificial intelligence, complex systems, agent-based modeling and simulation.

Contact: Marco Pistore pistore [at] fbk.eu

 C1 - Advancing Agricultural Sustainability through Data Intelligence (1 grant)

This initiative contributes to global food security and environmental sustainability, fostering a collaborative and innovative research atmosphere focused on integrating data science and artificial intelligence (AI) with agriculture. It addresses urgent challenges such as resource management, crop yield optimization, and the impacts of climate change. As the global demand for food increases, this scholariship promotes the adoption and creation of technologies and methodologies to boost agricultural efficiency and sustainability. The successful candidate will engage in developing innovative, data-driven methods for sustainable agricultural practices. Additionally, he/she will explore interdisciplinary research, utilizing remote sensing, machine learning, and predictive analytics to produce actionable insights aimed at enhancing agricultural productivity and resilience. This scholarship aims to cultivate a new breed of researcher, adept at leveraging data intelligence for agricultural progress, with a focus on optimizing water usage, enhancing crop resilience, and implementing sustainable farming practices.

Contact: Massimo Vecchio mvecchio [at] fbk.eu

C2 - Formal methods for embedded software (1 grant)

Techniques based on formal methods for the verification and validation of embedded and safety-critical software systems are becoming increasingly important, due to the growing complexity and importance of such systems in every aspect of modern society. Despite the major progress seen in the last twenty years, however, the application of formal methods in embedded software remains a challenge in practice, due to factors such as the interplay between computation and physical aspects and the increasing complexity of the software and its configurations.  
 This project will investigate novel techniques for the application of formal methods to the design, verification, and validation of embedded software, with particular emphasis on safety-critical application domains such as railways, automotive, avionics, and aerospace. The techniques considered will include a combination of automated and interactive theorem proving, satisfiability modulo theories, model checking, abstract interpretation, and deductive verification. Examples of the problems tackled during the project include the formal verification of functional requirements expressed in temporal logics, automated test-case generation, efficient handling of parametric/multi-configuration software systems and product lines, and the verification of software operating in a physical environment, subject to real-time constraints. Importantly, in addition to researching novel theoretical results, a significant part of the project activities will be devoted to the implementation of the techniques in state-of-the-art verification tools developed at FBK and their application to real-world problems in collaboration with our industrial partners.

Contact: Alberto Griggio griggio [at] fbk.eu

C3 - Formal methods for industry (1 grant)

Industrial systems are reaching an unprecedented degree of complexity. The process of designing a complex system is expensive, time consuming and error-prone. Moreover, the design process has to guarantee not only the functional correctness of the implemented system, but also its dependability and resilience with respect to run-time faults. Hence, the design process must characterize the likelihood of faults, mitigate possible failures, and assess the effectiveness of the adopted mitigation measures.  
Formal methods have been increasingly used over the last decades to deal with the shortcomings of designing a complex system. Formal methods are based on the adoption of a formal, mathematical model of the system, shared between all actors involved in the system design, and on a tool-supported methodology to aid all the steps of the design, from the definition of the architecture down to the final implementation in HW and SW. Formal methods include technologies such as model checking, an automatic technique to symbolically and exhaustively analyze all possible executions of the system in the formal model, in order to detect design flaws as early as possible. Model checking techniques have been recently extended to assess the safety and dependability characteristics of the design, and for system certification. 
The objective of this study is to advance the state-of-the-art in system design using formal methods. This includes adapting and extending the system design methodology, investigating improved versions of state-of-the-art routines for verification and safety assessment of complex systems, and developing novel extensions to address open problems. Examples of such extensions include novel techniques for contract-based design and contract-based safety assessment, advanced techniques for formal verification based on compositional reasoning, the analysis of the timing aspects of fault propagation, the characterization of transient and sporadic faults, the analysis of the effectiveness of fault mitigation measures in presence of complex fault patterns, and the modeling of analysis of systems with continuous and hybrid dynamics. 
This study will exploit the challenges and benchmarks defined in various industrial projects carried out at FBK.

Contact: Marco Bozzano bozzano [at] fbk.eu Alberto Griggio griggio [at] fbk.eu

C4 - Formal Methods for Digital Twins (1 grant)

Digital twins are dynamic and self-evolving models that simulate a physical asset and represent its exact state through bi-directional data assimilation. They employ Artificial Intelligence (AI) data-driven and symbolic techniques to provide state synchronization, monitoring, control and decision support. 
This project will build on the results of the ongoing ESA-funded project ExploDTwin (""Digital Twin for Space Exploration Assets""), which aims at integrating the DT into the ESA infrastructure to support online space assets operations with functionalities such as planning, what-if-analysis, fault detection, diagnosis and prognosis. To this end, ExploDT introduces a model-based design methodology that allows to seamlessly integrate engineering methods and AI techniques into a cohesive DT framework. 
The PhD student will investigate new formal methods to analyze the DTs by integrating model checking, automated theorem proving, simulation, and machine learning. Different aspects of the DTs will be considered including temporal properties for validation, monitorability and diagnosability. The new methods will be implemented and evaluated on space related benchmarks derived from ExploDTwin results.

Contact: Stefano Tonetta tonettas [at] fbk.eu

D1 - Advanced radar sounder data processing and information extraction (project ASI-INAF n. 2023-6-HH.0 JUICE-RIME-E — CUP no. F83C23000070005) (1 grant)

In the context of the European Space Agency (ESA) mission JUpiter ICy moons Explorer (JUICE) to the Jovian system, and the development of the ESA EnVision mission to Venus, we seek for candidates willing do develop methodologies for radar sounder data processing (data enhancement, semantic segmentation, denoising, content based retrieval, target detection, multitemporal analysis, etc.). The outcome of this activity will contribute in improving the understanding of the subsurface of planetary bodies, the correlation to history and climate as well as to a better understanding of the Earth.
The candidate will be requested to design and develop novel methodologies within artificial intelligence framework (machine learning, deep learning, pattern recognition, etc.) for effective information extraction from radar sounder data.
Besides the requirements established by the rules of the ICT school, preferential characteristics for candidates for this scholarship are:
•    master degree in Electrical Engineering, Communication Engineering, Computer/Data Science, Mathematics or equivalents;
•    knowledge in artificial intelligence, image/signal processing, remote sensing, radar remote sensing.
This scholarship is funded by project ASI-INAF n. 2023-6-HH.0 JUICE-RIME-E - “Missione JUICE - Attività dei team scientifici dei Payload per Lancio, commissioning, operazioni e analisi dati” — CUP no. F83C23000070005

Contact: Francesca Bovolo bovolo [at] fbk.eu

Additional position supported by a scholarship under the Italian National Recovery and Resilience Plan (NRRP) Mission 4, component  2

B2 - Securing Container and Microservices (BaC SERICS S4 SecCO, PE00000007; CUP D33C22001300002) (1 grant)

Containers emerged as a lightweight alternative to virtual machines that offer better microservice architecture support. There are many scenarios such as remote monitoring of critical infrastructures, emergency responses, traffic management and planning, etc. where moving functionality towards the edges has several benefits. Containers technology recognized this by developing container- orchestration systems (i.e., microk8s and k3s) specifically for edge devices. These benefits are even more emphasized if applications running on containers are developed as a combination of microservices. Scalability, portability and fault-tolerance are just few of the attributes that can be better exploited running microservices at the edge. On the other hand, security is more challenging because there is not a single point of enforcement, data and computations are distributed and the isolation properties are weaker in containers compared to more traditional virtual machines. Container security involves protecting containerized applications and their infrastructure throughout their lifecycle, from development to deployment and runtime. security should provide solutions to address the following threat landscapes: (I) protect a container from applications inside it, (II) provide inter- container protection, (III) protect the host from containers, and (IV) protect containers from a malicious or semi-honest host. An open research challenge is the design of solutions that can at the same time guarantee enforcement of runtime policies (i.e., control flow integrity) and acceptable overheads to be able to run on a wide range of edge devices. Also scalable and flexible access control techniques that accomodate the various infrastructural scenarios on which micro services can run is an open problem. Research activities that will be carried out within this PhD may include:

• Security Policies Alignment: Interaction with existing security policy specification language and contribute to the constructs to express properties and aspects of the security policies that need to be enforced at run-time compared to static enforcement and compliance performed at development time.
• Threat model and approach definitions: Specification of the threat model. Identification of the appropriate methodologies and technologies to enforce security policies at runtime.
• Novel approaches to run-time monitoring and anomaly detection. Design of new algorithms and methods that are scalable, and resistant to adversarial learning attacks.
• Mitigation approaches: Microservices are becoming popular also in safety-critical scenarios, thus it is of paramount importance to adopt non-disruptive mitigation approaches when a security policy breach is detected that limit attacks without stopping the system operations.

This activity studies in-depth how to design mitigations that are compatible with the safety and security constraints of the system.

Contact: Bruno Crispo bruno.crispo [at] unitn.it

GLASSFORM.ai SpA

A9 - AI-based Digital Twins and Control Systems for hollow glass manufacturing (1 grant) [additional scholarship (published on 05.08.20234] 

Within the context of hollow glass industrial manufacturing, this PhD research is aimed at exploiting physics-aware Artificial Intelligence solutions, including Physics Informed Neural Networks (PINNs) as well as Computer Vision and Time Series Forecasting, both with the aim of building Digital Twins and Simulators for the process as well as in conjunction with meta-heuristic optimizations such as multi-objective and bio-inspired algorithms, to minimize the environmental impact of hollow glass production processes while maximizing the quality of the final products.
The candidate will explore interdisciplinary research involving the development of Artificial Intelligence and Data-Driven solutions, optimization methods and continuum mechanics.
This initiative is funded by GlassFORM.ai, an international joint-venture between Bottero (Italy) and Tiama (France), worldwide leaders in machinery and inspection systems for glass production. GlassFORM.ai provides process automation solutions for the manufacturing industry using Artificial Intelligence and Machine Learning; its target market is the packaging industry, with main focus on hollow the glass production segment (bottles, containers, etc.).
Hollow glass production occurs by melting glass' constitutive matter at about 1700°C and then carrying out a wealth of processing steps some of which are described by continuum fluid dynamics, some other by discrete physics.
The entire process entails heat transfer, either from or towards glass, in order to trigger variations of state and behavior, and is characterized by a strong nonlinearity due to both glass' intrinsic properties as well as its interaction with the machinery.
Besides the requirements of the Doctoral School, the successful candidate:

• has a Master Degree in Mathematics, Physics, Engineering or related fields;
• has good understanding of Partial Differential Equations (PDEs) and their applications;;
• is familiar with Continuum Mechanics, especially in Fluid Dynamics and Heat Transfer;
• is proficient in Artificial Intelligence and Machine Learning;
• has either some degree of programming expertise, preferably in Python including core AI/ML libraries (PyTorch/Tensorflow, Scikits-Learn) or proficiency in MatLab/SimuLink.

The intellectual property rights pertaining to the research outcomes derived from the doctoral student's activities shall be vested in the company. Nevertheless, the moral rights of the doctoral candidate shall be respected.

Contact: Federico Monegaglia federico.monegaglia [at] glassform.ai